U.S. patent application number 12/252379 was filed with the patent office on 2009-12-31 for method and system for determining position of a wireless electronic device within a volume.
Invention is credited to Jacques Y. Guigne, Nicholas G. Pace, James A. Stacey.
Application Number | 20090325598 12/252379 |
Document ID | / |
Family ID | 41448076 |
Filed Date | 2009-12-31 |
United States Patent
Application |
20090325598 |
Kind Code |
A1 |
Guigne; Jacques Y. ; et
al. |
December 31, 2009 |
METHOD AND SYSTEM FOR DETERMINING POSITION OF A WIRELESS ELECTRONIC
DEVICE WITHIN A VOLUME
Abstract
A method for determining a position of a wireless electronic
device within a volume includes detecting a signal transmitted by
the wireless device during two-way communication to and from a
first known position within the volume. The method further includes
detecting the signal from at least three additional known positions
within the volume, where the at least three additional known
positions are spatially independent of each other. The method
further includes determining a phase difference between the signal
detected at the first position and the signal detected at each of
the at least three additional positions, determining the position
of the wireless electronic device using the phase differences, and
at least one of displaying and storing the position of the wireless
electronic device.
Inventors: |
Guigne; Jacques Y.;
(Paradise, CA) ; Stacey; James A.; (Paradise,
CA) ; Pace; Nicholas G.; (Bath, GB) |
Correspondence
Address: |
RICHARD A. FAGIN
P.O. BOX 1247
RICHMOND
TX
77406-1247
US
|
Family ID: |
41448076 |
Appl. No.: |
12/252379 |
Filed: |
October 16, 2008 |
Related U.S. Patent Documents
|
|
|
|
|
|
Application
Number |
Filing Date |
Patent Number |
|
|
61076702 |
Jun 30, 2008 |
|
|
|
Current U.S.
Class: |
455/456.1 |
Current CPC
Class: |
G01S 5/06 20130101; G01S
5/021 20130101 |
Class at
Publication: |
455/456.1 |
International
Class: |
H04W 64/00 20090101
H04W064/00 |
Claims
1. A method for determining a position of a wireless electronic
device within a volume, comprising: detecting a signal transmitted
by the wireless electronic device for two-way communication from a
first known position within the volume; detecting the signal from
at least three additional known positions within the volume, the at
least three additional known positions being spatially independent
of each other; determining a phase difference between the signal
detected at the first position and the signal detected at each of
the at least three additional positions; determining the position
of the wireless electronic device using the phase differences; and
at least one of storing and displaying the position of the wireless
electronic device.
2. The method of claim 1 wherein the transmitted signal is a
Bluetooth two-way communication signal.
3. The method of claim 1 wherein the volume is disposed within a
building.
4. The method of claim 1 wherein the position is determined in
three dimensions.
5. The method of claim 1 wherein the wireless electronic device
comprises a cellular telephone.
6. The method of claim 1 further comprising detecting the signal
from at least one additional known position within the volume.
7. The method of claim 1 wherein the first known position is
disposed approximately in the center of the volume.
8. The method of claim 1 wherein the phase difference is related to
a frequency of the detected signals to determine at least one time
difference between the detected signals.
9. An apparatus for determining a position of a wireless electronic
device within a volume, comprising: a reference receiver configured
to detect a signal transmitted by the wireless electronic device
and used as a corresponding reference signal; at least three
additional receivers which are spatially independent of each other,
said additional receivers being coupled to the reference receiver
to receive the reference signal and being configured to detect the
signal transmitted by the wireless electronic device; and a
processor configured to compute the position of the wireless
electronic device based on a phase difference between the reference
signal generated by the reference receiver and the signals detected
at each of the additional receivers.
10. The apparatus of claim 9, wherein the reference receiver is
configured to detect a Bluetooth two-way communication signal.
11. The apparatus of claim 9, wherein each of the additional
receivers comprises a phase detector circuit which produces an
output that is indicative of the phase difference between the
reference signal generated by the reference receiver and the signal
detected at each of the additional receivers.
12. The apparatus of claim 9 wherein the processor is configured to
determine a time difference between detection of the reference
signal and the signals detected at at least one of the additional
receivers based on a frequency of the detected signals.
Description
CROSS-REFERENCE TO RELATED APPLICATIONS
[0001] Priority is claimed from U.S. provisional application No.
61/076,702, filed on Jun. 30, 2008.
FIELD
[0002] The invention relates generally to the field of wireless
communication. More specifically, the invention relates to a method
and system for locating a wireless electronic device.
BACKGROUND
[0003] The ability to measure the position of a mobile electronic
device opens up a wide range of new applications. Applications that
depend on location include location-based services, location-based
advertising, context-aware computing, enhanced position
determination (e.g., global positioning system--"GPS"), enhanced
"911" and similar emergency response services, asset tracking and
real-time location services, autonomous robotic systems, advanced
man-machine interfaces, and assistive technologies for the
disabled.
[0004] The availability of mobile electronic technology has
substantially increased over the past decade. Cell phones, for
example, have essentially saturated the market in the most
developed countries. The proliferation of cell phones has
introduced problems in the delivery of emergency services such as
responding to "911" emergency calls due to the fact that cell
phones are difficult to locate geodetically with sufficient
accuracy. The U.S. Federal Communications Commission ("FCC") has
mandated that enhanced 911 services be supported by the cellular
telephone infrastructure, but the providers of cell phone services
are limited by the technology capability of their networks, and are
presently able to locate individual cell phones to within a range
of about 50-300 meters of the actual phone location. In a dense
urban environment, such precision is insufficient to properly
locate a 911 caller. Despite the crude precision of the cell phone
positioning, social networking has emerged as a new location-based
service delivering revenue for the cell phone service
providers.
[0005] The commercialization of the United States Armed Services
GPS allowed this infrastructure to support the positioning of
mobile electronic devices by calculating the distance between the
device and at least four low earth orbit satellites in the
constellation of GPS satellites. The emergence of handheld GPS
receivers created applications for GPS location services. The key
limitation of GPS is that it is unable to deliver position
information inside buildings. Enhanced GPS has emerged as a new
technology direction as innovators seek to extend the functionality
of GPS to the indoors. This has created the need for new
infrastructure known as indoor, or in-building, positioning
systems.
[0006] In the Enterprise Resource Planning sector, a technology
known as Radio Frequency Identification, or RFID, has become an
essential part of supply chain management systems that promote the
tracking of inventory and assets in the business. The idea behind
RFID is that relatively cheap, "smart" tags are used to identify
goods and/or assets and a sophisticated RFID system is used to
locate and identify the smart tags for tracking purposes. For many
companies, location-aware technology is conferring competitive
advantage in the marketplace.
[0007] One of the newest markets for location aware services is in
the area of online advertising. Companies such as Google, Inc.,
Yahoo, Inc. and Microsoft, Inc. are aggressively competing in this
field. By introducing information concerning target recipient
location into online advertising, the value of an online ad can be
greatly increased. Location based advertising promises to become
the next front in the battle for online advertising market
share.
[0008] Context aware computing is yet another technology that has
been promoted by the large computing manufacturers and stimulated a
great deal of R&D in computer science, engineering, and
industry. As computing becomes increasingly mobile, the day will
come when all computing will become location aware.
[0009] Market analysts have been predicting that location based
services will be a multi-billion dollar market, but the timing of
this prediction has been problematic. The problem is that, as
technology becomes increasingly pervasive, consumers are becoming
more aware and concerned about issues such as privacy. If there is
any piece of personal information that would be viewed by the
consumer as sensitive, the ability of an all-pervasive technology
to track an individual's position would rank near the top of the
list of concerns.
[0010] Contemporary suppliers of location based services have a
wide range of technology on which such services are based.
Positioning technologies use detectors based on light or sound.
Technologies using light include optical detectors, radio frequency
detectors, and infrared detectors, which are all special cases
using the electromagnetic spectrum. Sound waves can also be used
with systems using ultrasound at frequencies outside the range of
human hearing above 20 KHz.
[0011] Technology suppliers have been attempting to enable improved
context-aware computing by improving the resolution of location
measurement systems. The state-of-the-art at present seems to be
location resolution in the range of 1-10 meters. Expensive systems
are available that can locate to fractions of a centimeter, but
such systems cannot be deployed as part of a more pervasive
location aware infrastructure.
[0012] In addition to the technology used (light or sound), there
is also the methodology employed to determine position that
determines the effectiveness of the method. Two methods of
determining position common in the present market are time (or time
difference) of arrival techniques ("TOA" and "TDOA") and received
signal strength indication ("RSSI"). TOA and TDOA allow calculating
the position of a mobile electronic device by measuring the range
from a transmitter to a multiplicity of receivers using timing
electronics and knowledge of the speed of electromagnetic energy
(or sound) through air. By determining the range between the
transmitter and at least three independently positioned receivers,
the three dimensional position may be calculated using
trilateration. In the more general problem of locating a mobile
device, GPS uses multilateration to calculate the position of a
mobile GPS receiver. Four satellites are needed because there are 4
unknowns in the GPS problem, three values for the position (X, Y,
Z) and one value for time.
[0013] For indoor position determination, TDOA systems require a
multitude of receivers scattered through the surveillance volume.
The cost of such systems is relatively higher as at least three
receivers must be within range of the client device (the device to
be located) and such receivers need to be networked together,
independently powered, individually calibrated, etc. Deploying such
infrastructure is cumbersome because overhead costs scale with
respect to the number of receivers.
[0014] Another positioning system that is being deployed uses RSSI.
Technologies such as Wireless Local Area Networks ("WLAN" or
"WiFi") and Bluetooth have RSSI built in. The idea is that the
proximity of a client device to a WLAN device can be inferred from
signal strength of radio transmissions between the client and WLAN
device. Using complex algorithms and learning networks, the rough
position of the client can be inferred. The advantage of such
systems is that the coverage of existing WLAN and WiFi network hubs
is quite high and the incremental cost of implementing a
positioning system on RSSI is very low. The disadvantage is that it
doesn't make much sense to increase WLAN penetration beyond what is
needed to provide basic connectivity. The accuracy of RSSI is not
much better than 10 meters, or "room level."
[0015] The ubiquity of location servers will be limited until the
cost of individual location servers becomes as cheap as other mass
market consumer devices. Further, it is not only the cost of the
location servers that must be taken into account. The cost of
implementing the corresponding client location hardware and
software that will limit the adoption rate of this technology must
also be considered. Issues such as the cost of deploying the
location aware infrastructure, the cost of maintaining and
calibrating the infrastructure and the delivery of value-added
services on that infrastructure will all play a role in the growth
of this market.
[0016] However, it is evident that the present resolution of
location aware devices is not sufficient to fully enable or deliver
the promise of context aware computing. A breakthrough technology
is required with a resolution that is an order of magnitude better
than the current state of the art. One such technology that is
being positioned as potentially delivering new levels of accuracy
is Ultra-Wideband ("UWB"), which promises resolutions of order 15
cm-1 m with a multiplicity of receivers approach. Basically, the
UWB system uses very narrow pulses to increase the resolving power
of the TOA/TDOA approach. In order to shape a very narrow pulse,
very large bandwidth is required.
[0017] US Patent Application Publication No. 2006/0199534 A1
(Location System for Bluetooth Enabled Devices) by Smith describes
a method, apparatus and system for tracking and locating Bluetooth
enabled devices. In this application, a network of Bluetooth
sniffers is used to locate "lost" devices and their owners. A
"parent" device independently monitors received signal strength
between itself and a "child" device. When the signal strength of
the child drops below a certain level, the child is deemed by the
parent to be lost and an alert is issued to the sniffer network by
the parent. Upon receipt of the alert, the sniffer network is then
engaged to locate the child device through paging for the child
device throughout a network of Bluetooth capable sniffer devices.
The location is determined by proximity to a particular sniffer
device at a known location. The method of the Smith '534
publication provides room level resolution in locating a child
device which is adequate for this "lost and found" application. The
methodology used uses RSSI (received signal strength indication) as
the underlying technology.
[0018] U.S. Pat. No. 6,819,286 B2 (Location Determination For
Mobile Units) issued to Armbruster et al. describes a method for
location determination using Bluetooth techniques within buildings,
underground or within other structures. The method disclosed in the
'286 patent uses a multiplicity of subsidiary units arranged in a
geometric pattern within the surveillance volume. A minimum of
three subsidiary units are needed to measure the range to a mobile
device to determine its position through trilateration techniques.
This method is essentially the time delay of arrival method and
relies on timing circuits for its implementation. In addition, the
subsidiary units are each independently deployed through the
surveillance volume and must be individually powered and networked
together. The overall accuracy of the method is strongly dependent
on the latencies of communications between the mobile unit and each
of the subsidiary units.
[0019] U.S. Pat. No. 6,745,038 B2 (Intra-Piconet Location
Determination and Tomography) issued to Callaway et al. describes a
novel technique for intra-piconet location determination and
tomography using received signal strength indication (RSSI). In
this invention, the range between two piconet devices is determined
by analyzing the destructive interference between direct and
reflected wavepaths. By examining the RSSI versus carrier-frequency
curve and determining the frequency separation of the nulls, the
range may be determined. In principle, the method is capable of
determining the range with an accuracy between 2.62 cm and 1 meter.
When the position of a reflector is located at the origin, a system
of equations describing the relative ranges between devices in the
piconet can be solved to determine the positions of each device. In
two dimensions, range measurements between a minimum of five
devices are needed for solution while, in three dimensions, seven
independent range measurements are needed.
[0020] U.S. Pat. No. 6,717,516 (Hybrid Bluetooth/RFID Based Real
Time Location Tracking) issued to Bridgelall describes a hybrid
device that allows RFID tags to be identified and located using
Bluetooth technology. A plurality of fixed devices is distributed
over an area containing the items to be tracked. The fixed devices
are operated as RFID readers to identify and locate items having
RFID tags. The fixed devices are preferably distributed at
distances corresponding to twice the range of the devices when
operated as RFID readers. The location of the RFID tag is inferred
by several methods. The first method cited is locating a mobile
slave device to within a piconet cell by determining which of the
fixed devices is associated with the slave device. The resolution
of this method is 30 feet (10 meters), equal to the maximum range
of a Class 2 Bluetooth device. A second method is described whereby
the Bluetooth cell size is adjusted to equal the range of the RFID
passive tag reading capability, which has a resolution of 12-15
feet (anticipated to increase in the future with advancing
technology). Finally, in a particularly preferred example, range
may be determined from the phase of the response signals and the
phase may be determined at a plurality of frequencies to resolve
phase ambiguities. The problem of locating the position of a mobile
device is addressed through defining a directional antenna beam
pattern to limit the RFID tag exposure to a narrow cone angle. By
changing the beam direction through electronic or mechanical beam
steering the angular position of the RFID tag may be
determined.
[0021] US Patent Application Publication No. 2002/0180640 (Location
Estimation in Narrow Bandwith Wireless Communications Systems) by
Gilkes et al. uses the phase difference between a known stable
reference signal and a known signal output by a plurality of
wireless mobile communications devices (location markers) at
several known locations to determine the location of a mobile
wireless communications device transmitting in the ISM radio band.
The phase of the 1 MHz signal transmitted by the mobile device
allows the phase difference to be detected within the location
estimation environment within 300 meters (the wavelength of the 1
MHz signal). The location markers measure the phase difference
between the embedded signal (the 1 MHz bitstream output by the
Bluetooth radio) and a 1 MHz sine wave frequency reference signal
that is produced at a fixed location by a stationary reference
oscillator and is distributed to the location markers by coaxial
cable, modified Ethernet or latency-free wireless means. A system
calibration procedure is required to determine a phase delay
parameter that measures the propagation delay between the reference
source and each location marker. A 1 MHz phase comparator measures
phase to 0.001 cycles (6.2832 milliradians), yielding 30 cm
resolution (11.8 inches) in the range measured between the location
marker and the mobile device. The location solution processor uses
information from at least 4 non-coplanar location markers and
solves simultaneous equations derived from the Cartesian
coordinates of the location markers and the differences between the
relative times of arrival reported by the location markers.
SUMMARY
[0022] In one aspect, the invention relates to a method for
determining a position of a wireless electronic device within a
volume which comprises detecting a signal transmitted by the
wireless electronic device at a first known position within the
volume. The method includes detecting the signal from at least
three additional known positions within the volume, the at least
three additional known positions being spatially independent of
each other. The method includes determining a phase difference
between the signal detected at the first known position and the
signal detected at each of the at least three additional positions.
The method includes determining the position of the wireless
electronic device using the phase differences and at least one of
storing and displaying the position of the wireless electronic
device.
[0023] In another aspect, the invention relates to an apparatus for
determining a position of a wireless electronic device within a
volume which comprises a reference receiver configured to detect a
signal transmitted by the wireless electronic device, this signal
acting as a reference signal. The apparatus includes at least three
additional receivers which are spatially independent of each other.
The additional receivers are coupled to the reference to receive
the reference signal and configured to detect the signal
transmitted by the wireless electronic device. The apparatus
further includes a processor configured to compute the position of
the wireless electronic device based on phase differences between
the reference signal generated by the reference receiver and the
signals detected at each of the additional receivers.
[0024] Other features and advantages of the invention will be
apparent from the following description and the appended
claims.
BRIEF DESCRIPTION OF DRAWINGS
[0025] The accompanying drawings, described below, illustrate
typical examples of the invention and are not to be considered
limiting of the scope of the invention, for the invention may admit
to other equally effective examples. The figures are not
necessarily to scale, and certain features and certain views of the
figures may be shown exaggerated in scale or in schematic in the
interest of clarity and conciseness.
[0026] FIG. 1 depicts a system for determining a position of a
wireless electronic device.
[0027] FIG. 2 is a schematic of a Bluetooth radio.
[0028] FIG. 3 is a schematic of a Bluetooth radio with phase
difference array.
[0029] FIG. 4 depicts a 3D phase difference array geometry with
three receiver pairings.
[0030] FIG. 5 depicts a 3D phase difference array geometry with
four receiver pairings.
[0031] FIG. 6 depicts a geometry of multi-lateration equations.
[0032] FIGS. 7A-7D are graphs depicting positioning accuracy of
phase difference array.
[0033] FIGS. 8A and 8B depict unwrapping of phase at phase-wrapping
boundary.
DETAILED DESCRIPTION
[0034] The invention will now be described in detail with reference
to a few examples, as illustrated in the accompanying drawings. In
describing the examples, numerous specific details are set forth in
order to provide a thorough understanding of the invention.
However, it will be apparent to one skilled in the art that the
invention may be practiced without some or all of these specific
details. In other instances, well-known features and/or process
steps have not been described in detail so as not to unnecessarily
obscure the invention. In addition, like or identical reference
numerals are used to identify common or similar elements.
[0035] FIG. 1 depicts a system 101 for determining the position of
a wireless electronic device within a volume. Different
combinations of components and interfaces define different systems,
depending on the final application. In one example, the system 101
includes the following components: a client device (CLI-1) 100, a
transmitter or transceiver (TX1) 102, a receiving sensor array
(RX-ARRAY) 104, a receiving base station (BASE-1) 106, and a
backend server (SERV-1) 108. The client device 100 is a wireless
electronic device whose position is being measured. The client
device 100 may be mobile or stationary, handheld or not. The
transceiver 102 transmits an RF pulse under command of a user
through the client device 100 (client mode) and/or under the
command of the receiving base station 106 (server mode). The client
device 100 and transceiver 102 may be physically integrated into a
single device or provided as separate devices with an appropriate
communication link between them. The client device 100 may receive
signals from the receiving base station 106 through a receiver
(RX1) 103 or transceiver 102.
[0036] The receiving sensor array 104 includes RF sensors (not
shown separately) in a geometric array. The RF sensors enable the
calculation of the three-dimensional (3D) position of the client
device 100 by detecting RF signals issued by the transceiver 102.
The receiving base station 106 receives electronic signals from the
receiving sensor array 104 and employs digital signal processing
techniques to process the received signals on multiple sensors and
then calculates the 3D position of the client device 100. The
receiving base station 106 reports this position to the client
device 100 or the backend server 108, where the position may be
communicated to and displayed by the client device or processed and
communicated to the client as part of a value-added service by the
server. The receiving base station 106 and receiving sensor array
104 may be integrated into a single device or may be provided as
separate devices with an appropriate communication link between
them. It is the collocation of the receiving elements within the
receive array that is salient to the present invention. In
practice, network devices are connected together over a LAN or WAN.
Key components of the LAN are network hubs that define the network
topology. The receiving base station 106 may include a wireless hub
to facilitate communication with client device(s).
[0037] The system 101 can communicate with the World Wide Web (WWW)
110, for example, in order to provide location or value-added
services to the client device 100. The backend server 108 acts as a
portal to WWW 110 (and/or a communications network) and may deliver
location services to the client device 100 (primarily in the event
that the WWW is unavailable). The client device 100, backend server
108, and receiving base station 106 may communicate with the WWW
110 and a "cloud" of location aware services to which the client
device 100 has subscribed.
[0038] User 111 can interact with the client device 100 through
user interface (UIF-1) 112. Users of client devices may be
subscribers to location-based devices. User 113 can interact with
the backend server 108 through a restricted user interface (UIF-2)
114. Users of the backend server 108 may be restricted to system
support personnel. User 115 can interact with WWW 110 through
restricted user interface (UIF-3) 116. Users of WWW 110 that can
gain access to the system 101 and the client device 100 may be
restricted to internet service providers (ISPs). Client device 100
communicates with transceiver 102 through digital interface (IF-1)
118. The primary function of the interface 118 is to command the
transceiver 102 to transmit data. Client device 100 may also
communicate with receiver (RX-1) 103 through an interface 118a
similar to digital interface 118. The function of the transmitter
102 and receiver 103 may be combined in a single device using a
transmit/receive switch to enable bi-directional digital
communications 124. Transmit and receive functions in the client
device are separated in FIG. 1 as the signals transmitted by TX-1
need not be the same signals as those signals used to establish
two-way communications between the base unit and the client
electronic device. An air interface (IF-2) 120 is provided between
the transceiver 102 and the receiving sensor array 104. The air
interface 120 represents the transport of an RF signal in air
between the transceiver 102 and the receiving sensor array 104.
[0039] The receiving sensor array 104 communicates with the
receiving base station 106 via an analog interface (IF-3) 122. Each
sensing element in the receiving sensor array 104 will generate
analog electrical signals corresponding to the RF signals received
from the client device 100. The client device 100 communicates with
the receiving base station 106 through a bi-directional digital
interface (IF-4) 124, as noted above. The bi-directional digital
interface 124 supports the delivery of location services to the
client device 100 (absolute or relative position with respect to
the receiving base station 106) independently of the availability
of the backend server 108 or access to WWW 110. In one or more
examples of the invention, the bi-directional interface 124 uses
the Bluetooth communications protocol. The receiving base station
106 communicates with the backend server 108 through a
bi-directional digital interface (IF-5) 126. The backend server 108
may be configured with software to deliver value-added location
based services to the client device 100 through industry standard
interface (SOA-2) 128 in the event that WWW 110 is unavailable. The
bi-directional digital interface 126 will typically conform to the
Internet (TCP/IP) and WWW (XML/HTTP) protocols. The receiving base
station 106 communicates with WWW 110 through a bi-directional
digital interface (IF-6) 130. WWW 110 may also be configured with
software to deliver value-added location based services to the
client device 100. The bi-directional digital interface 130
supports the ability to update the receiving base station 106
software and provide access to the receiving base station 106 (for
remote monitoring and control).
[0040] Web services are delivered between WWW 110 and the client
device 100 independently of system 101 via the industry standard
interface (SOA-1) 132. SOA is an acronym that denotes a service
oriented architecture. In general, industry standard interface 132
does not traverse the firewall 134 that separates the system 101
from WWW 110. Web services are delivered between the backend server
108 and client device 100 through the industry standard interface
(SOA-2) 128. These services may be a subset or superset of the
services the client device 100 might access via the interface 132.
Interface 128 acts as a backup interface to interface 132 for those
services delivered from WWW 110 and provides the value-added
services that may be unique to system 101 (i.e.,
application-dependent services). Services between WWW 110 and
backend server 108 are delivered via industry standard interface
(SOA-3) 136. Interface 136 allows the backend server 108 to be a
proxy for WWW 110 and allows remote access to the backend server
108 for monitoring, control, and maintenance via interface 116.
[0041] The present invention can be implemented using commercially
available RF technology (analog or digital) at low cost. A
practical example is described using Bluetooth radio technology for
indoor positioning applications up to 10 m in range (corresponding
to Class 2 Bluetooth technology). Within the range of the proposed
device, all Bluetooth devices communicate within a so-called
"piconet." Multiple position server devices extend the capability
of positioning within wider surveillance volumes by deploying these
devices as a so-called "scatter-net."
[0042] Bluetooth radio is attractive as a technology for the
present invention as it supports a high level communications
protocol to facilitate the communication of value added services
between location server devices and client devices. Bluetooth
separates the 2.4 GHz Industry, Scientific and Medical (ISM) band
into 79 channels of 1 MHz bandwidth. This channel separation
provides the Bluetooth device with the ability to isolate itself
from other devices transmitting in this band by allowing the
Bluetooth device to transmit across each of the available channels
by using a pseudo-random "hop" sequence. The Bluetooth
specification supports data packets that can be retransmitted if
errors in transmission occur.
[0043] The range of carrier frequencies supported by the Bluetooth
standard allows the present invention to calculate phase
differences between Bluetooth receiver pairs as a function of
carrier frequency and thereby calculate the position of a Bluetooth
client to within a few centimeters. These phase differences may be
determined using commercial off-the-shelf analogue RF components.
Alternative examples using digital components are equally feasible
as discussed below.
[0044] The characteristics of Bluetooth radio are as follows:
[0045] Bluetooth operates in the 2.4 GHz band. In the US and
Europe, a band of 83.5 MHz width is available; in this band, 79 RF
channels spaced 1 MHz apart are defined. In France, a smaller band
is available; in this band, 23 RF channels spaced 1 MHz apart are
defined. [0046] The wavelength of electromagnetic waves at 2.4 GHz
is 0.125 m. [0047] The communication channel is facilitated by a
hopping sequence hopping through the 79 or 23 RF channels. Two or
more Bluetooth devices using the same channel form a piconet. There
is one master and one or more slave(s) in each piconet. The hopping
sequence is unique for the piconet and is determined by the
Bluetooth device address of the master. [0048] The channel is
divided into time slots where each slot corresponds to an RF hop
frequency. Consecutive hops correspond to different RF hop
frequencies. [0049] The channel is divided into time slots, each
625 us in length. The time slots are numbered according to the
Bluetooth clock of the piconet master. [0050] A TDD (time division
duplex) scheme is used where master and slave alternatively
transmit. The master starts its transmission in even numbered time
slots only, and the slave shall start its transmission in
odd-numbered time slots only. [0051] All data on the piconet
channel is conveyed in packets. Each packet consists of 3 entities,
the access code (68/72 bits), the header (54 bits), and the payload
(0-2745 bits). [0052] The prescribed hop rate is 1600 hop/second so
that the duration spanning each hop is 625 us. [0053] The Bluetooth
radio module uses GFSK (Gaussian Frequency Shift Keying) where a
binary one is represented by a positive frequency deviation and a
binary zero by a negative frequency deviation. The shape of the
Gaussian filter (through the BT parameter) is set to 0.5 and the
modulation index must be between 0.28 and 0.35. [0054] Gaussian
Frequency-Shift Keying (GFSK) is a type of Frequency Shift Keying
modulation that utilizes a Gaussian filter to smooth
positive/negative frequency deviations, which represent a binary 1
or 0 where the minimum deviation is 115 kHz and the maximum
deviation is 175 kHz.
[0055] FIG. 2 is a schematic of one embodiment of a Bluetooth radio
200 (known as a superheterodyne radio). The Bluetooth radio 200
corresponds to or is a component of the receiving base station (106
in FIG. 1). The Bluetooth radio 200 consists of a baseband
processor 202, transmitter 204, receiver 206 (called Receiver 0),
and antenna 208. The baseband processor 202 contains a
microprocessor (CPU) 210 with random access memory (RAM) 212 and
flash memory (ROM) 214 which can be shared with a digital signal
processor (DSP) 216. The baseband processor 202 controls the
Bluetooth radio 200. In one example, the baseband processor 202
controls the switching between channels and performs the time
division duplex (TDD) control.
[0056] On the transmit side of the radio, digital data at 1
Mbit/sec is converted to an analog bit stream using a
digital-to-analog converter (DAC) 217 and passed through a Gaussian
low pass filter (LPF) 218 to eliminate the high frequency
components that would leak outside the desired channel. A FM
modulator 220 modulates an intermediate frequency (IF) carrier and
up-converts the modulated signal to the final radio frequency (RF)
corresponding to the desired channel. The desired channel is
selected through a frequency hop control circuit 222 containing a
crystal reference oscillator (not shown), a phase-locked-loop (PLL)
224, a loop filter 226, and a voltage controlled oscillator (VCO)
228. The output of the FM modulator 220 is amplified by a power
amplifier 230 and switched to the antenna 208 on transmit (Tx) on
even numbered time slots of the master Bluetooth device. On both
transmit and receive the signals are passed through a band-pass
filter 232 with bandwidth of 84 MHz centered on 2442 MHz.
[0057] On the receive side of the radio, a low-noise amplifier
(LNA) 234 provides about 10-15 dB of gain prior to the mixer 236.
The frequency hop control circuit 222 provides a local oscillator
frequency corresponding to the desired 1 MHz bandwidth channel
through the local oscillator frequency LO output from the voltage
controlled oscillator (VCO) 228. The channel frequency of the
received harmonic signal is at frequency RF, which is defined for
the k-th channel as
RF=.omega..sub.k+h*MOD(t) (1)
where .omega..sup.k is the angular frequency of the k-th channel
(equal to 2.pi.f.sub.k where f.sub.k=(2402+k) MHz, k=0,78), h is
the modulation index, and MOD(t) is the modulation function that
defines the frequency modulation of the Bluetooth bit stream. In a
standard Bluetooth radio, the RF signal is mixed with the LO signal
to down-convert the signal to an intermediate carrier frequency IF.
The modulation and phase of the RF signal is unchanged by this
down-conversion. An FM demodulation circuit (FM Demod) 238 then
infers the frequency modulation of the signal and outputs the
envelope of the demodulated signal. This demodulated signal is then
interpreted by an analog-to-digital (ADC) converter 240 as a
bit-stream of 0s and 1s. After the FM demodulation, all phase
information in the original carrier signal is lost.
[0058] Referring to FIG. 3, it is on the receive side of the
Bluetooth radio 200 that the present invention advances the current
state of the art. The present invention implements a phase
difference array (PDA) by preserving the absolute phase of the
original RF frequency in the 2.4 GHz band and calculating the phase
differences between pairs of geometrically independent receivers
(in a manner that is described below). First, it is shown that it
is possible to preserve the phase information of the original
signal when the signal is down-converted to an intermediate
frequency IF. Next, it is shown that the down-converted RF signal
from one antenna may be used as the reference signal for all
remaining signals from an array of geometrically independent
antennas. The signal from each antenna is then processed by an
independent radio receiver. Finally, by pairing the signals from
geometrically independent antennas relative to the reference
antenna in the manner indicated herein, the difference in phase
between the original signals in the 2.4 GHz band that correspond to
the time difference of arrival may be measured with a novel new
approach that does not require timing electronics. These
measurements are digitized and passed back to the baseband
processor 202 to calculate the position using a phase difference
algorithm that solves the non-linear trilateration equations.
[0059] In FIG. 3, an array of receivers 242 (corresponding to the
receiving sensor array 104 in FIG. 1) communicates with the
Bluetooth radio 200. In the illustrated example, the array of
receivers 242 includes receiver 206a (called Receiver 1), receiver
206b (called Receiver 2), and receiver 206c (called Receiver 3). In
general, the array of receivers 242 may include three or more
receivers. Let .phi..sub.rf be the absolute phase of the RF signal,
.phi..sub.lo the absolute phase of the local oscillator,
.phi..sub.1 the phase of the transmitted signal at Receiver 1 and
.phi..sub.0 the phase of the transmitted signal at Receiver 0.
Receiver 0 and Receiver 1 represent one pairing of receivers in the
proposed phase difference array comprising the present invention.
Other receiver pairings would be Receiver 2 and Receiver 0,
Receiver 3 and Receiver 0, and so on. In this arrangement, the
Receiver 0 is the reference receiver with which all receiver
pairings are made. The summed total of these receiver pairings
comprise the receiving phase difference array. If the antenna for
Receiver 0 is taken to be at the origin of the array, then the
positions of the remaining antennas define the geometry of the
array. The remaining antennas are arranged such that they are
spatially independent.
[0060] The mixing of the signals between the receivers in a
receiver pairing defines the phase difference that is unique to the
present invention. Mixing of signals will now be described for
receiver pairing Receiver 0 and Receiver 1, but the same principle
can be applied to any receiver pairing in the system, e.g.,
Receiver 0 and Receiver 2, Receiver 0 and Receiver 3, and so on.
Let S.sub.1(t) be the harmonic RF signal received at Receiver 1 and
S.sub.0(t) be the harmonic RF signal received at Receiver 0
(reference receiver). Further, let R(t) be the reference signal
generated by the local voltage controlled oscillator (VCO) 228 at
frequency LO=.omega..sub.k+.DELTA..omega., where .omega..sub.k is
the angular frequency of the k-th channel and .DELTA..omega. is a
constant angular frequency offset (typical values for
.DELTA..omega. are 110.6 MHz, 110 MHz, or 43 MHz). Then,
S.sub.1 (t)=V.sub.1 cos(RF*t+.phi..sub.1+.phi..sub.rf) (2)
S.sub.0 (t)=V.sub.0 cos(RF*t+.phi..sub.0+.phi..sub.rf) (3)
R(t)=V.sub.r cos(LO*t+.phi..sub.lo) (4)
where the amplitudes of the harmonic and VCO outputs are V.sub.1,
V.sub.0 and V.sub.r, respectively.
[0061] The common reference signal R(t) output by the VCO 228 of
Receiver 0 is routed to the mixer electronics 236a, 236 for both
Receiver 1 and Receiver 0, respectively (and similarly for each
receiver pairing in the array). Each mixer in the receiver array
outputs signals (in this case, MIX.sub.1(t) and MIX.sub.0(t) are
the mixer outputs for Receiver 1 and Receiver 0, respectively) that
combine the harmonic signals as follows
MIX 1 ( t ) = LPF ( R ( t ) * S 1 ( t ) ) = LPF ( V r cos ( LO * t
+ .PHI. lo ) * V 1 cos ( RF * t + .PHI. 1 + .PHI. rf ) ) ( 5 ) MIX
0 ( t ) = LPF ( R ( t ) * S 0 ( t ) ) = LPF ( V r cos ( LO * t +
.PHI. lo ) * V 0 cos ( RF * t + .PHI. 0 + .PHI. rf ) ) ( 6 )
##EQU00001##
[0062] The mixer combines the harmonic signals with the reference
signal and outputs sum and difference terms for the combined
signals, which can be calculated using a well-known mathematical
identity for the product of cosine functions to be
R ( t ) * S 1 ( t ) = ( 1 2 K 1 V r V 1 ) [ cos ( ( LO - RF ) * t +
( .PHI. lo - .PHI. rf ) - .PHI. 1 ) + cos ( ( LO + RF ) * t + (
.PHI. lo + .PHI. rf ) + .PHI. 1 ) ) ] ( 7 ) R ( t ) * S 0 ( t ) = (
1 2 K 0 V r V 0 ) [ cos ( ( LO - RF ) * t + ( .PHI. lo - .PHI. rf )
- .PHI. 0 ) + cos ( ( LO + RF ) * t + ( .PHI. lo + .PHI. rf ) +
.PHI. 0 ) ) ] ( 8 ) ##EQU00002##
where K.sub.1 and K.sub.0 have dimensions [1/V] and represent the
mixer gains. The low pass filter function LPF(*) filters out the
high frequency component, leaving the following as the final output
of the mixers:
MIX 1 ( t ) = ( 1 2 K 1 V r V 1 ) cos ( ( LO - RF ) * t + ( .PHI.
lo - .PHI. rf ) - .PHI. 1 ) ( 9 ) MIX 0 ( t ) = ( 1 2 K 0 V r V 0 )
cos ( ( LO - RF ) * t + ( .PHI. lo - .PHI. rf ) - .PHI. 0 ) ( 10 )
##EQU00003##
[0063] The outputs of the mixers 236, 236a are then passed through
bandpass filters 260, 260a, respectively, centered on the
intermediate frequency IF=LO-RF=.DELTA..omega. with a bandwidth of
1 MHz. For the other receiver pairs, similar outputs of mixers
236b, 236c would be passed through bandpass filters 260b, 260c,
respectively, centered on the intermediate frequency
IF=LO-RF=.DELTA..omega. with a bandwidth of 1 MHz. A Bluetooth
radio may have multiple stages of mixing, amplification and
filtering at successive intermediate frequencies, but the output of
each such IF-stage is similar--the output being a new intermediate
frequency carrier with a modulation signal, a constant phase
difference (.phi..sub.lo-.phi..sub.rf) and a unique receiver phase
that passes through the mixer unchanged.
[0064] At this point, the outputs of the mixers (236, 236a) are
processed differently. The output MIX.sub.0(t) of mixer 236 is
passed to the FM demodulation circuit 238 of the Bluetooth Receiver
0 to extract the modulation signal and output the bitstream
corresponding to the digital data of the radio signal. In this way,
Receiver 0 can act as a normal Bluetooth device (master or slave)
and process digital communications signals according to the
Bluetooth protocols. The output of the FM demodulator circuit 238
is converted from an analog voltage to a digital output by the
analog-to-digital converter (ADC) 240 and routed back to the
baseband processor 202.
[0065] Simultaneously, the output MIX.sub.0(t) of mixer 236 is
passed to the receiver array 242 electronics to serve as the phase
reference for the phase difference array. In the illustrated
example, the outputs MIX.sub.1(t) from mixer 236a and MIX.sub.0(t)
from mixer 236 are passed to a phase detector circuit 270a (which
will be comprised of phase comparators, filters and associated
electronics) that outputs a voltage that is proportional to the
phase difference of the signals MIX.sub.1(t) and MIX.sub.0(t). This
output is simply (.phi..sub.0-.phi..sub.1), the difference in
absolute phase due to the path length difference in the transmitted
signal to Receiver 0 and Receiver 1, respectively. All other
components of the signal (due to the carrier frequency, the
modulation of the carrier frequency, and the absolute phases of the
radio frequency and oscillator signals) cancel out, leaving an
output voltage representing the phase difference of the receiver
pair. Similar outputs will be output by the other receiver pairs at
phase detector circuits 270b, 270c. The outputs of the phase
detector circuits 270a, 270b, 270c are then converted from an
analog voltage to a digital output by analog-to-digital converters
(ADC) 240a, 240b, 240c, respectively, and routed back to the
baseband processor 202, which determines the location of the
transmitting client device (100 in FIG. 1) using an algorithm
described in more detail below.
[0066] In the present invention, the phase reference for the entire
system is taken to be the RF signal received on one of the radio
receivers of the receiver array. Thus, when the phase differences
corresponding to time differences of arrival (TDOA) of the RF
signal across each receiver pairing of the receiver array are
processed, the constant phase differences between the LO and RF
signal that do not correspond to a time difference of arrival at
each receiver all cancel out. The remaining phase difference due
only to the TDOA across each receiver pair (where each receiver in
the array is paired with the reference receiver which is taken to
be located at the origin) is then the phase difference used to
calculate the position of the client device that is transmitting at
that time.
[0067] There are only a finite number of places in the analog
receiver processing chain where the signal can be tapped to provide
the signals for the methods used by the present invention. In FIG.
3, the point at which the signal is extracted from the radio
circuit is indicated just before the FM demodulator 238.
Alternative examples are possible and are described below.
[0068] The circuitry to analyze the phase difference between
receiver pairs can be implemented with analog phase detectors. The
output of a phase detector is typically a current sink/source that
can be used to drain or charge the capacitor of a low-pass filter
(effectively acting as an integrating circuit) to yield a voltage
output that is a known function of the phase difference on the
range (-pi,+pi). This circuit is very similar to the PLL 224, Loop
Filter 226, VCO 228 combination employed in the frequency hop
control circuit 222, but in this case, it is the voltage output
from the integrating circuit that is desired. With current
technology, phase differences can be calculated with fractional
errors of 1e-3 to 1e-4.
[0069] The analog circuitry for the receiver array 242 is indicated
in FIG. 3. Note that, in this simplified schematic, many
intermediate stages in the radio (amplification and filtering) are
not shown. Note also that the figure only shows three receivers
paired with the reference receiver (Receiver 0). Three receiver
pairs is a minimum configuration for 3D location measurements.
[0070] The Bluetooth radio 200 connected to the origin antenna 208
is a full-featured Bluetooth radio (as described above) with
baseband processor 202, transmitter 204, and receiver 206 and acts
as the controller for the entire system and communicates with
external Bluetooth devices in its reception volume. This radio
provides the local oscillator frequency reference LO from the
channel frequency control circuit 222 that switches all receivers
to the same channel. Receiver 0 also provides the analog reference
signal MIX.sub.0(t) that is compared with the signals received on
each of the receiver antennas (208a, 208b, 208c) in the array of
receivers 242. Each receiver signal is compared with the reference
signal and a phase difference is calculated by a phase detector
circuit (270a, 270b, 270c). It is the analog voltage corresponding
to the phase difference from each receiver pair that is digitized
by an ADC (240a, 240b, 240c) and passed to the DSP 216 in the
baseband processor 202. Ideally, the voltage output from the phase
detector (270a, 270b, 270c) is linearly related to the phase
difference. However, once the output is digitized, any
non-linearities can be corrected by the baseband processing
algorithms.
[0071] The fact that Receiver 0 corresponds to a normal Bluetooth
radio allows the error correction circuitry of the baseband
processor 202 to be employed to validate the phase difference data
output by the phase difference array. The baseband processor 202
can correlate digital data from Bluetooth packets with the
digitized output of the phase difference array containing those
packets in real-time. If a Bluetooth communications packet is
received without errors, then it can be assumed that the carrier
signal for that packet has been received without interference (say
from multipath reflections) and the phase differences based on that
uncorrupted packet can also be presumed to be uncorrupted by
interference.
[0072] Although the proposed example outlined above is based on
analog electronic components, the output of the mixers (236, 236a,
236b, 236c) can be superposed on an essentially arbitrary
intermediate frequency IF without losing the phase information. In
an alternative example of the proposed invention, the output of the
mixers could be passed directly through ADCs and the output stored
in memory for subsequent digital signal processing to determine the
phase differences directly in the digital domain using known
methods in the art of digital signal processing.
[0073] In another alternative example of the proposed invention,
the output of the low-noise amplifiers at each of the receiver
antennas (208, 208a, 208b, 208c) could be digitized just prior to
the mixer electronics (236, 236a, 236b, 236c) and the entire radio
implemented digitally. In each such example, the processing chain
of the phase difference array as described above may be implemented
using the art of digital signal processing. As the transfer
functions of each of the analog components are known in the art of
analog radio electronics, there is no theoretical impediment to
implementing any of the analog components described above as an
equivalent digital processing algorithm. The engineering choice
between analog versus digital processing of the signals from the
phase difference array in a given example will be driven primarily
by the availability of low cost commercial off-the-shelf (COTS)
components from which a commercially viable device can be
constructed.
[0074] The algorithms that implement the Phase Difference Array
(PDA) method have the following novel features: [0075] A method of
accurate position determination using narrowband (<1 MHz
bandwidth) radio signals over a range of carrier frequencies in the
Industrial, Scientific and Medical (ISM) band at 2.4 GHz. [0076] A
method of calculating the time difference of arrival of multiple
signals to high accuracy without a timing circuit or common time
reference. [0077] A method of improving the accuracy of position
determination using long continuous wave (CW) signals used by
existing radio technologies independently of the frequency
modulation of the carrier. [0078] A method of improving the
accuracy of position determination by defining the payload of a
special communications packet when the client radio issues a
position location request. [0079] A method of improving the
accuracy of position determination by specifying the channel
frequency hop sequence subsequent to a client position location
request.
[0080] The time difference of arrival of multiple signals can be
calculated to high accuracy without a timing circuit or common time
reference. This phase difference array method identifies the time
difference of arrival as being equal to the slope of phase
differences between RF receiver pairs as a function of RF carrier
frequency. This is a new application of a known method in the art
of sonar signal processing. One of the advantages of this approach
is that this method can be applied to arrays of receiver pairs
separated by greater than 1/2 wavelength without phase
ambiguities.
[0081] The position of the client RF device is then calculated by
solving the trilateration equations for arrays of 3 receiver pairs
and solving the multi-lateration equations for arrays with more
than 3 receiver pairs. The attainable accuracy for determining
position advances the current state of the art (typically 1 m
accuracy for WiFi and RFID) by 1-2 orders of magnitude. The
apparatus and methods in this invention processes positions with
accuracies in the centimeter range. This is an important innovation
in narrowband RF position determination and opens up new
applications and market opportunities.
[0082] Another attractive feature of the proposed approach allows
position to be determined using narrowband RF signals without
explicit calculations of the range. Typical RF ranging is strongly
dependent on accurate timing and discrimination of RF pulses, with
the accuracy of the method dependent on the pulse width. Narrow
pulse widths require a very broad spectrum of frequencies
(broadband signals). An example of this approach is ultra-wideband
(UWB) ranging. This approach is infeasible with narrow-band signals
employed widely today. The ability of measuring position with
narrow-band signals allows the present invention to calculate
position of a communications channel during transmission. This
allows the system to combine position measurements with data
communications. Special packets can be defined with payload data
specified that enhances the ability of the system to calculate
position.
[0083] Often in short baseline devices it is the phase difference
alone that is used to obtain angles and the range is obtained by
other means. In such cases special considerations are needed to
have both a baseline of a few wavelengths and a means of resolving
the 2.pi. ambiguity that arises for baselines longer than a half
wavelength. In the present case it is important to note that the
preferred method actually determines (X, Y, Z) directly without
explicitly calculating the range or angles. The time delays are
obtained to sufficient accuracy by measuring the phase difference
between sensors at a number of the Bluetooth hop frequencies. As
noted above, the slope of the line of phase differences versus
frequency gives the time delay directly.
[0084] The GFSK modulated signals received by the antennas are
processed such that they occupy a frequency band 84 MHz wide in the
2.4 GHz band. The frequency shift to define bits as 1 or 0 is about
115 kHz. A packet of data contains about 2745 bits after which
another carrier frequency is used from the 79 channels
available.
[0085] The phase difference between two sensors is measured over
the duration of a single packet which by request may have been
assigned a location mode in which all bits are either 1 or 0 (or a
pattern of alternating series of 0s and 1s such that the filtering
effects of the Gaussian filter are negligible). The accuracy of the
phase difference measurement is significantly enhanced over the
direct output of the phase comparator circuit as it is obtained
over a time period of about 600 .mu.secs which corresponds to many
thousands of cycles.
[0086] During the next received packet the phase difference is
again measured, but it is now at a different hop frequency. This
continues over a range of the hop frequencies. If during the
location mode all the possible hop frequencies were used the time
required would be 79*2*625 .mu.sec which corresponds to about
1/10.sup.th sec. But in practice a smaller range would be used.
Perhaps only the 23 hop frequencies available in France might be
used.
[0087] As the rate of change of phase with frequency is time, so
the slope of the phase difference between sensors with frequency is
the time delay between the sensors. If the phase difference between
sensors separated by a baseline of N wavelengths becomes greater
than 2.pi. as the frequency varies then an unwrapping of the phase
difference would be necessary. This is possible but preferably to
be avoided. The phase wrapping problem is illustrated in FIGS. 8A
and 8B. Plotting the phase difference versus frequency to obtain
the time delay it may be necessary to make one adjustment of 2.pi.,
as illustrated. Using robust estimation techniques, for example,
the necessary adjustment for phase unwrapping can be implemented in
software.
[0088] The slope of a straight line through the data is calculated
in a least squares manner and provides the estimate of the time
delay. If each point on the phase difference versus frequency plot
has a standard deviation of a, then the standard deviation of the
time delay estimate is
.DELTA. t = 12 a 2 / ( N h ( N h 2 - 1 ) ) 1 f c ( 11 )
##EQU00004##
where N.sub.h is the number of hop frequencies used and where a is
the error in the measurement of phase difference in radians.
[0089] The actual time between sensors is
t = N f c ( 12 ) ##EQU00005##
where f.sub.c the mean carrier frequency and N is the spacing in
wavelengths.
[0090] The fractional error in the time delay measurement as a
function of the number of hop frequencies and the baseline in
wavelengths can then be written as
.DELTA. t t = ( 12 a 2 N h ( N h 2 - 1 ) ) 1 N ( 13 )
##EQU00006##
This relationship shows that the accuracy of the method described
herein depends on three complementary techniques: 1) increasing the
accuracy of the phase detection circuitry (relative errors of
0.001-0.0001 with current technology), 2) increasing the number of
hop frequencies N.sub.h over which the phase slope is calculated,
and 3) increasing the baseline of the phase difference array.
[0091] Referring to FIG. 6, the idea behind all the
multi-lateration equations solved herein is that a transmitter
emits a spherical wave S originating at its true position
(X.sub.0,Y.sub.0,Z.sub.0) and which propagates to the 3D phase
difference array at the speed of light c. The signals received on
each element of the phase difference array will have absolute
phases relative to the phase of the signal received at the origin
of the array. By pairing the receivers relative to the origin,
receiver pairings are defined and through these pairings, phase
differences may be defined. From these phase differences the time
differences of arrival are determined as described above. From
these time differences, path differences may be defined simply by
multiplying each time difference by c. With the path differences
between the sensors and the wavefront defined, the geometry of the
path differences relative to that spherical wavefront is determined
and defines the geometry of the multi-lateration problem.
[0092] It is important to note that the methods of this invention
apply to phase difference arrays with a geometry that can be
defined very generally, with different separations (i.e. baselines)
for individual receivers, different orientations of the receiver
pairs, different numbers of receiver pairs, and different numbers
of reference antennas defining independent phase difference arrays,
all within a single compact device. The specific examples of phase
difference array geometries for three and four receiver pairs
outlined below illustrate how the method can be applied to these
cases. Application to the general case follows in a straightforward
fashion.
[0093] In FIG. 4, a 3-dimensional phase difference array with 3
receiver pairs is defined with 4 antennas (0, 1, 2, 3) at positions
(0,0,0), (1,0,0), (0,1,0), and (0,0,1) where all positions are in
units of s, the separation of the antennas from the origin. If each
antenna is separated from the origin by one wavelength, then s is
12.5 cm at 2.4 GHz. The antenna at the origin provides the
reference signal for calculating the phase differences at the
antennas located at the unit vectors along the x, y and z axes. The
phase differences correspond to path length differences d.sub.1,
d.sub.2 and d.sub.3 for the x, y and z antennas respectively. For
simplicity, all calculations are expressed in units of s. The
methods described here may be extended in a straightforward fashion
to more general phase difference array geometries where the
antennas are located at different separations and with different
orientations relative to the origin.
[0094] The nonlinear equations to be solved find (X, Y, Z) such
that
(X-1).sup.2+Y.sup.2+Z.sup.2-(D-d.sub.1).sup.2=0 (14)
X.sup.2+(Y-1).sup.2+Z.sup.2-(D-d.sub.2).sup.2=0 (15)
X.sup.2+Y.sup.2+(Z-1).sup.2-(D-d.sub.3).sup.2=0 (16)
where
D.sup.2=X.sup.2+Y.sup.2+Z.sup.2 (17)
These equations represent 4 equations in 4 unknowns. Substituting
for D, these nonlinear equations can be simplified to
f 1 = X - d 1 X 2 + Y 2 + Z 2 - 1 2 ( 1 - d 1 2 ) = 0 ( 18 ) f 2 =
Y - d 2 X 2 + Y 2 + Z 2 - 1 2 ( 1 - d 2 2 ) = 0 ( 19 ) f 3 = Z - d
3 X 2 + Y 2 + Z 2 - 1 2 ( 1 - d 3 2 ) = 0 ( 20 ) ##EQU00007##
In this form, the equations are straightforward to solve for
(X,Y,Z) using the Newton-Raphson method in three dimensions, for
example. An exact solution may also be derived.
[0095] In the present invention there is no dependence on timing
circuits to determine the range. The position may be obtained
directly from the path differences obtained from the phase delays
alone. Additionally, there is no dependence in the method on
measurements of angles. The phase difference array described here
does not use beam steering to determine position. The position is
determined by solving the exact equations for a spherical wave
emitted from a point source. The method is thus an exact method
that provides accurate position measurements in the near-field
(traditional phased arrays typically are formulated with far-field
approximations to these exact equations and require range and angle
measurements to recover position).
[0096] As noted above, the accuracy of the method can be further
improved by improving the resolution of the phase difference
circuit, increasing the separation of the antennas, and increasing
the number of hop frequencies. Additionally, it is possible to
increase the number of cycles used to determine the phase, and
adding additional receiver pairs (or adding additional arrays). The
case of adding an additional receiver pair is now examined.
[0097] The effect of adding an additional receiver pair into the
phase difference array is not only to improve the overall accuracy
of the method, but also to allow the multi-lateration equations to
be linearized and solved analytically. In FIG. 5, a 3D phase
difference array with 4 receiver pairs is defined with 5 antennas
(0, 1, 2, 3, 4) at positions (0,0,0), (1,0,0), (-a, -b, 0), (-a, b,
0) and (0,0,1) where all positions are in units of s, the
separation of the antennas from the origin, a=cos(30) and
b=cos(60). We can then define
d 1 = D - ( X - 1 ) 2 + Y 2 + Z 2 ( 21 ) d 2 = D - ( X + a ) 2 + (
Y - b ) 2 + Z 2 ( 22 ) d 3 = D - ( X + a ) 2 + ( Y + b ) 2 + Z 2 (
23 ) d 4 = D - X 2 + Y 2 + ( Z - 1 ) 2 ( 24 ) ##EQU00008##
The following six equations can be constructed by taking sensors 1,
2, 3 two at a time where the .alpha..sub.i and .beta..sub.i can be
expressed in terms of the d.sub.i and a and b (and implicitly
s):
2X=.alpha..sub.1D+.beta..sub.1 (25)
2X=.alpha..sub.6D+.beta..sub.6 (26)
2Y=.alpha..sub.2D+.beta..sub.2 (27)
2Y=.alpha..sub.3D+.beta..sub.3 (28)
2Y=.alpha..sub.5D+.beta..sub.5 (29)
2Z=.alpha..sup.4D+.beta..sub.4 (30)
where 2.beta..sub.5=.beta..sub.3+.beta..sub.2 and
2.alpha..sub.5=.alpha..sub.3+.alpha..sub.2. Thus three linear
equations in D can be found
D 16 = .beta. 1 - .beta. 6 .alpha. 6 - .alpha. 1 ( 31 ) D 35 =
.beta. 3 - .beta. 5 .alpha. 5 - .alpha. 3 ( 32 ) D 52 = .beta. 5 -
.beta. 2 .alpha. 2 - .alpha. 5 ( 33 ) ##EQU00009##
[0098] Any of the foregoing equations can be used directly for an
estimate of D, and then (X, Y, Z) can be obtained from the above
equations. Each of the equations contains d.sub.1, d.sub.2, and
d.sub.3 and provides the same estimate. Simulations have shown that
even with very large phase errors the three solutions give
effectively the same result.
[0099] The non-linearity of the tri-lateration equations can be
removed by the additional receiver pair that, in effect, solves for
the range D directly (D is the source of non-linearity in
determining X, Y, Z). The additional receiver pair provides
increased robustness as the equations can be solved exactly and
provides increased accuracy.
[0100] The accuracy of the phase difference of arrival algorithm is
tabulated in Table 1 below and illustrated in FIGS. 7A-7D. For the
case where the baseline is one wavelength (the nominal baseline)
and the relative error of the phase detector circuit is 1e-3 (the
conservative case), and calculating the phase differences at 50
different hop frequencies, the standard deviation of the error in
calculating range is 2.25 cm in the plane defined by |X|.ltoreq.15
m, |Y|.ltoreq.15 m, and Z=2 m. FIG. 7A shows the differences
between the mean range predicted and true range in cm in this plane
has a standard deviation of 2.2503 cm with a mean absolute
difference of 1.5214 cm (when the standard phase error is 0.001
radians, number of frequency hops is 50, and the baseline is 1
wavelength). FIG. 7B shows the differences between the mean X
predicted and true X in cm has a standard deviation of 1.5607 cm
with a mean absolute difference of 0.98693 cm. FIG. 7C shows the
differences between the mean Y predicted and true Y in cm has a
standard deviation of 1.5919 cm with a mean absolute difference of
0.99156 cm. FIG. 7D shows the differences between the mean Z
predicted and true Z in cm has a standard deviation of 0.21975
cm.
TABLE-US-00001 TABLE 1 Std Dev of differences between true and
extracted range for Phase error Baseline in Z |X| .ltoreq. 15 m,
|Y| .ltoreq. 15 m (radians) wavelengths (m) (cm) 0.001 1 2 2.25
0.0001 1 2 0.23 0.0001 0.5 2 0.93 0.0001 0.5 1 0.93 0.001 0.5 1
5.8
[0101] The manufacturing of the phase difference array introduces
errors in the positions of the antennas that, when combined with
the finite size of the antennas, necessitates a calibration to
determine the "phase centers" of the antennas. This calibration can
be performed after the base unit is manufactured and before the
unit is shipped. During calibration the base station is positioned
at a number of known positions relative to a transmitter and the
time delays measured. The predicted and actual transmitter position
is then brought into coincidence using well-known methods in
nonlinear optimization in several variables to give the positions
of the phase centers of each sensor in the array.
[0102] In Table 2 below, the results of a simulation of sensor
spacing recovery using the proposed calibration method is presented
(all spacing of receivers is nominally 1 wavelength, or 125 mm).
The data shows that the phase centers of the antennas may be
determined to high accuracy (<0.1 mm) when the number of
transmitter positions used in the calibration is equal to 100. This
indicates that the phase difference array approach may be
calibrated to achieve the stated levels of accuracy inherent in the
method.
TABLE-US-00002 TABLE 2 Error in transmitter 1.5 cm 3 cm 6 cm coords
at X = Y = 15 m, Z = 2 m Phase error (radians) 0.001 0.001 0.001
Initial assumed 125 mm 125 mm 125 mm value of s Real value of s 130
mm 130 mm 130 mm Mean value of final s 130 mm 130.00 mm 129.99 mm
Std. Dev. of error 0.076 mm 0.099 mm 0.12 mm Error in predicting
1746.37 mm 1758.92 mm 1752.4 mm X at X = 15 m due to error in s
with original value of s Error in predicting 25.86 mm 33.92 mm
40.20 mm X at X = 15 m due to std. dev. in final value of s
[0103] In the present invention, a solution for real-time location
determination is based on narrowband radio frequency (RF)
technology and uses a novel new apparatus and methods collectively
embodied as a "phase difference array". The proposed technology and
methodology have a competitive advantage in the accuracy of
position that can be rendered versus competing technologies (<5
cm) and by virtue of this accuracy enables new classes of
applications that can exploit this accuracy (e.g., location based
advertising at point of sale, assistive technologies, indoor
navigation systems, context aware computing). The only technology
that appears to offer similar accuracy in the market segments of
interest is ultra-wideband (accuracy about15 cm). Other
technologies are progressively worse in overall accuracy: RFID
(.about.1 m), WLAN (.about.1m), GPS (.about.10 m outdoors, no
accuracy indoors), cell phone location, E911 (50 m-300 m). Optical
locating systems have potentially very high accuracy (<1 cm) but
at high cost and limited applicability (primarily robotic systems).
Competitive ultrasonic systems provide only room level
accuracy.
[0104] A practical example of the proposed invention is based on
Bluetooth radio as the underlying technology. Bluetooth is a very
widely deployed technology that uses the RF spectrum in the
Industry, Scientific and Medical band at 2.4 GHz. Using Bluetooth
allows the proposed location serving technology, method and system
described herein to be used with the 100 s of millions of Bluetooth
enabled devices already in the market. The proposed system takes
advantage of the economies of scale that have already been achieved
by Bluetooth devices. The proposed device provides accurate
position determination in a small compact device. It will be
substantially easier to deploy and calibrate in the field. The
proposed invention can also be manufactured using commercial
off-the-shelf hardware. The proposed invention is also designed to
integrate easily into a "cloud" of intelligent location aware
infrastructure delivering value added services to mobile electronic
devices.
[0105] The present invention has been designed to address a much
more refined and diverse range of applications than the device
proposed by Smith (U.S. Patent Application Publication No.
2006/0199534 A1). In a very real sense, the present invention is an
enabling technology for a broad range of new applications. Where
the device described by Smith employs RSSI to determine position to
room level precision using existing capabilities of Bluetooth radio
technology, the invention described herein uses phase delays for
much higher resolution (order of centimeters versus tens of meters
in this example), which significantly advances the art of Bluetooth
technology.
[0106] The present invention promises location accuracies in the
centimeter range without the attendant complexity of the approach
of Callaway et al. (U.S. Pat. No. 6,745,038 B2). The present
invention uses direct measurement of phase differences to attain
its accuracy, while Callaway et al. employ RSSI and multipath
reflections and interference patterns to determine the range
between devices. The advantage of the present device is that only
two devices are required to determine the position of the client
device; namely, the client itself and the proposed location serving
technology of the present invention. The presence of 5-7 devices in
the piconet is not required.
[0107] We now examine the claims of Bridgelall (U.S. Pat. No.
6,717,516) and their applicability to the present invention. The
ability of translating phase differences into time differences of
arrival is well known in the art of digital signal processing. The
method of Bridgelall essentially duplicates the method of Gilkes et
al. (US Patent Application 2002/0180640), which has been addressed
in the background discussion. Time differences can, in turn, be
translated into a range between the transmitting device and the
receiving device providing that phase ambiguities can be resolved.
As no method for resolving phase ambiguities is presented by
Bridgelall, it must be assumed that the method used is part of the
prior art. In the present invention, we claim that the slope of the
phase differences as a function of frequency directly yields the
time differences of arrival without requiring timing circuits and,
most importantly, the method presented has no phase ambiguities to
be resolved. This technique, known previously in the art of sonar
signal processing, is claimed herein as an advance to the current
art of radio signal processing as applied to determining positions
of mobile electronic devices.
[0108] After the range is determined, Bridgelall claims only that
the position of the mobile device is determined as "an absolute
distance from (the fixed) device" with "high resolution". Such a
claim provides only a location for the mobile device as being
located somewhere on a sphere with a radius equal to the range
centered on the fixed device. It is only by introducing the well
known art of beam steering (electronically or mechanically) that
Bridgelall is able to refine the position of the mobile device by
determining angles of transmission between the mobile device and
the fixed device. In the present invention, a method is given that
determines to high accuracy the three-dimensional position X, Y,
and Z relative to a location server (i.e. the fixed device of
Bridgelall's method) without the need to determine either the range
or the angles directly. In addition, the present invention does not
use beam steering.
[0109] The high resolution claimed by Bridgelall is not specified
but it is evident that by calculating the range at independent
frequencies with independent measures of the phase differences
(i.e. translated into time differences of arrival), the method of
Bridgelall is attaining higher resolution by the known art of
statistical averaging of multiple independent measurements. In the
present invention, high accuracy is inherent in the method. Higher
accuracy is obtained not only by taking similar advantage of
statistical averaging over multiple frequencies but, in addition,
higher accuracy is obtained by calculating the phase difference
over many cycles of a continuous wave narrowband signal. Most
importantly, the narrowband signal used in the present invention
directly processes the phase differences of the carrier signals in
the ISM radio band (2.4 GHz) which is at a much higher frequency
than that used by Bridgelall (1 MHz). This latter frequency is the
frequency of the amplitude and phase output of the Bluetooth radio
after I/Q demodulation as defined by the Bluetooth standard. The
higher frequencies used by the present invention is an engineering
advance that has much higher inherent accuracy for location
determination.
[0110] The invention described herein encloses all the detectors
required to locate the mobile unit within a single compact unit for
reduced power and cost. The trilateration equations can be solved
to much higher precision if the time delay of arrivals can be
measured with high accuracy. This is achieved by the present
invention without requiring timing circuitry. The method used here
relies on phase differences rather than time delays directly.
Finally, the accuracy of the method proposed herein is greatly
improved over the method of Armbruster et al. (U.S. Pat. No.
6,819,286 B2) since the communications latencies are explicitly
cancelled out by locating the receivers in close proximity within a
single detector device.
[0111] The unique aspects of the present invention cited above in
examination of the claims of Bridgelall apply similarly to the
claims of Gilkes et al. In addition, the present invention does not
require a stable external reference oscillator to determine the
phase differences. It is a unique feature of the present invention
that no external reference signal requiring a complex and expensive
distribution network is required. In addition, the necessity that
the entire system be calibrated to determine phase delay constants
for each location marker makes the proposed system of Gilkes et al.
very difficult to set up and maintain. Anytime a location marker is
moved, the phase delay for that device would need to be
redetermined. The present invention also has much higher inherent
accuracy versus the method of Gilkes et al. by virtue of the
processing of the phase differences at the carrier frequencies in
the ISM band (2.4 GHz) versus the frequency of the Bluetooth radio
output (1 MHz). The system of Gilkes et al. is dependent on a
plurality of location markers positioned throughout the
surveillance volume to measure the position of a mobile device
whereas the present invention provides a stand-alone location
server contained within a single compact device that is much easier
to deploy. The overall accuracy of the method of Gilkes et al.
would evidently be no better than the 30 cm claimed for a given
range calculation while the present invention determines position
to <5 cm. This is the accuracy that is inherent in the unique
capability of the present invention to measure the phase difference
at the carrier frequencies in the 2.4 GHz ISM band.
[0112] While the invention has been described with reference to a
limited number of specific examples, those skilled in the art
having the benefit of the foregoing description will be able to
devise other examples that do not depart from the scope of the
invention as disclosed herein. Accordingly, the invention should be
limited in scope only by the attached claims.
* * * * *